In recent years, artificial intelligence has made significant progress in the field of medical diagnosis, especially in disease detection using facial recognition technology. Here are some major developments and applications:
-
A research team from Peking University developed an AI system that can detect chronic diseases such as hypertension, diabetes, and fatty liver by scanning the face with a thermal imaging device.
-
Skincare companies use facial recognition technology to assess skin conditions and recommend products, with accuracy approaching that of dermatologists.
-
FDNA's Face2Gene application can recognize over 300 rare genetic syndromes through facial features, with an accuracy rate of up to 85.7%.
-
A research team from the University of Melbourne developed a tool that can diagnose stroke by recognizing facial muscles, with an accuracy rate of 82%.
-
NEC (Nippon Electric Company) developed technology that can estimate pulse rate, oxygen levels, and respiratory rate through facial scanning in 10 seconds.
The principles behind these technologies are based on the following points:
- About 6,000 genes are associated with facial features, and gene mutations or deletions are reflected in facial characteristics.
- AI can perform precise quantitative analysis of facial features, identifying subtle changes that are difficult to detect with the naked eye.
- Temperature changes in different areas of the face may indicate certain diseases.
However, these AI technologies still face some challenges:
- They are mainly used for preliminary screening and cannot completely replace doctors' diagnoses.
- They require large amounts of high-quality patient data for training.
- Currently, most systems are trained on white patients, resulting in lower recognition accuracy for other races and skin colors.
Nevertheless, AI-assisted medical diagnostic technology continues to develop, and is expected to play an important role in broader disease detection and health monitoring in the future.